Explanations for Religious Diversity
Summary. I consider various hypotheses that are consistent with the existence of an eternal hell but that also explain why there has been such religious diversity throughout human history.
Conditional on the existence of an infinite hell, what probabilities should we assign to various hypotheses about it? This is what I call the "Pascalian approach" to religious questions (though Pascal himself would likely have demurred, as his original wager mentioned only heaven). The general method for solving such a question is clear enough: Start with a prior over the space of possibilities, and update it based on the evidence we have about the world. The details are, of course, more complicated.
In this piece, I'll explore the impact of conditionalizing on the fact that we observe a wide array of religious beliefs, including many held by cultures that have never been exposed to hell-proclaiming religions. This fact presents a problem for traditional notions that God both wants everyone to be saved and requires them to adopt certain beliefs or practices to do so. Below are some hypotheses that explain the observed facts. I haven't commented on the prior probabilites of the hypotheses, but I do think they're not all equal.
1. God does not want everyone to be saved.
Of the explanations I consider, this seems to me the one most widely held among traditional theologians. Examples of scriptural support for this view include Romans 9:14-24 and Qur'an 32:13, among many others. Indeed, "predestination" is one of the principal schools of religious thought throughout history.
In view of the fact that we appear to live in a deterministic universe, this hypothesis would seem to follow automatically from the assumption that God designed the universe: The initial conditions (plus God's interference) determine who will ultimately be saved, and it was up to God to determine the initial conditions the way he wanted. However, the inference to God's intentions about salvation also requires the assumption of divine foreknowledge: That God knew how things would turn out, which is the reason he designed them the way he did. The latter assumption is questionable, though. It seems to me that the easiest way for a being to attain knowledge of what will happen in our complicated universe of intricate cause-and-effect interactions would be to simulate it. But in that case, unless direct simulation doesn't actually produce consciousness, why couldn't we be part of that initial simulation? Or, for that matter, why not just conclude that God ran the real universe without needing to simulate it all ahead of time to peek at the answer?
So it seems reasonable that God might have established conditions for consignment to hell and set the universe in motion, without figuring out ahead of time who exactly would be damned. Rather than saying "God does not want everyone to be saved," then, we might say that "God set up the parameters for salvation, but didn't work to make sure that everyone would end up meeting them" (or even necessarily "coming close" to meeting them, as with, say, millennia of Native Americans who never heard of Jesus).
2. God did make an effort to save all cultures by revealing himself to them, but nearly all "went astray" and worshipped other gods.
Under a divine omniscience view, coupled with determinism, this would be indistinguishable from "God doesn't want everyone to be saved," because God would have known ahead of time that his efforts at revealing himself would not be enough. This needn't be true if God didn't peek at how things would turn out.
As a practical matter, hypothesis (2) isn't much different from (1). The difference lies only in the extent to which God tried to cause people to meet his salvation parameters by interfering with his world. Hypothesis (2) does explain why religion is nearly universal among human cultures, though it does seem odd that so many of them would have corrupted his original message (whatever that might have been) so badly.
3. God set different salvation requirements for different cultures, depending on what he revealed to them.
So, for instance, Jews in ancient Israel would have had to have followed Jewish laws to be saved, while contemporary Muslims in Iran would have to follow the Qur'an. This raises the question of what God expects of people who have heard more than one religious tradition, though.
4. Salvation is not based on beliefs that have not have been transmitted to you; God judges you based on your actions or "what's in your heart."
Hypothesis (3) could be seen a special case of this one, in which the required actions are "to follow the religious prescriptions of the culture in which you were raised." We can imagine other conditions that might be required, such as following a particular set of universal rules (e.g., don't murder others, don't steal, avoid pride, exercise self-moderation, etc.). The set of universal rules seems generally more likely to contain moral notions that are widely held among human societies, because otherwise one has to explain why God allowed so many cultures to develop moral sensibilities that were inconsistent with his will. So, for instance, punishment for dishonesty seems more likely than punishment for approaching ceremonial "water and fire during [...] menstruation" (Book of Arda Viraf, Ch. 20). (Perhaps. Though it is surprising how many religions have taboos related to menstruation.)
As far as Pascal's wager is concerned, hypothesis (4) doesn't necessarily imply contradictory actions with respect to the others. For instance, obeying a particular God often entails actions consistent with generally accepted notions of morality and spiritual discipline.
Hypothesis (4) has to account for why, if judgment is not based on beliefs, a large portion of the religious world thinks that it is. We could invoke an explanation in the manner of hypothesis (1): God doesn't want all to be saved and so fools some into thinking that "faith alone" is enough. But such an action would seem "unfair" to almost every culture on earth and so would seem inconsistent with a God who punishes according to widely shared moral values. (Of course, an eternal hell is also widely seen as unfair....) Another explanation is that sola fide notions arose naturally as corruptions of God's message, and once established, they remained highly persistent, because they sowed the seeds of their own memetic survival.
5. All religious notions about hell arose for reasons other than divine revelation. Punishment is based on something else entirely.
This category includes a large number of possibilities, including punishment based on something unexpected (e.g., Did your grandmother have brown hair?). However, unlike religious hypotheses, many of these scenarios have very little probabilistic interaction with the evidence we observe. No one has proposed that brown hair is more hell-worthy than black hair, so the probability of observing the world as we do on either hypothesis would seem to be exactly the same, or else so nearly the same that we needn't bother considering it. However, some non-religious hell scenarios are not entirely symmetric and deserve to be explored.
This class of scenarios must explain why people developed notions of eternal punishment but yet associated them with entirely the wrong sorts of criteria. A sadistic god who wants to punish people for random things would work, but why would he bother giving people suggestions about eternal punishment in unrelated contexts instead of hiding it from them entirely? Needless to say, there is little a Pascalian could do about such a god, except perhaps reduce the number of souls that come into existence. Even that isn't an obvious win, because a malevolent god might easily thwart such efforts. It's probably best to focus on scenarios other than those involving entirely arbitrary sadists.